Internals
Accompanies the specification. The spec states what a conforming implementation must do; this states how this one does it, and why the alternatives were rejected.
1 Scope · 2 Invariants · 4 Discovery · 5 Addressing · 8 Election · 9 Memory · 11 Wire · 12 Security · 16 Conformance · 17 Limits
Each names a different layer.
They are distinct and keyed differently.
The measured evidence for the second is that per-request cost is identical at 25 serial and 269 concurrent. If coalescing were doing the compression work, that figure would move.
Prior surface raises — assembly to C, manual memory to managed, threads to structured concurrency — removed a category of code from the application. Distributed applications still contain their communication: endpoints, retries, ordering, schema versions, and a list of consumers to notify.
The test is mechanical: take a system you maintain and count what remains if none of the message-handling code has to exist.
The simulator runs the real discovery, routing and election code against a modelled topology; this is possible because none of that code identifies the source of a value.
Moved here from the specification index. The specification defines FrogNet; it is indifferent to who or what reads it, and guidance about tooling does not belong in a document an independent implementer is meant to build from.
It is clear and obvious to the author of FrogNet that artificial intelligence has advanced sufficiently to assist in the development and deployment of large, complex systems such as FrogNet. It has not advanced sufficiently, nor does it possess the capacity for imagination and independent architectural judgment, to operate independently. It is strongly recommended that AI assistance be treated as a working engineering partnership between an architect and a junior- to mid-level programmer.
FrogNet is delivered with a complex, real-world simulator that exercises the live code across various topologies and under varying conditions. There are specific AI assistance files for the simulator that enable the generation of additional scenarios. We strongly recommend familiarity with the simulator and its capabilities, and that you incorporate it into both your manual and automated workflows.
The documentation set contains AI primers, architectural summaries, definitions, cross-references, and other embedded material specifically intended to give an AI assistant the context necessary to reason about FrogNet accurately.
When using an AI assistant with FrogNet, provide it access to the FrogNet documentation rather than relying on its general knowledge. FrogNet differs from conventional architectures in several areas, and an assistant without the supplied context may substitute conventional models that do not describe FrogNet.